Analysis:ThefutureweareenemiesorfriendswithAI?(TofreedownloadNews)Weareatacriticalperiodofartificialintelligenceresearch,artificialgeneralintelligencemaycomeatanytime.Thistechnologyoutofthelaboratoryhasbeguntohelptraditionalindustriesandthecontinueddevelopmentofmodernindustry.Butifonedaysomeonediesatthehandsoftherobothowtodo?Variousoutstandingissues,intheendtheeraofartificialintelligencewouldbeanothergoldenageofmankindorourother‘IceAge’?BillGates,AndyHerronMaske,Hawking,thesetechnologiesandtheoriesgiantshavewarnedusoftheemergenceofartificialintelligencecouldbeamajorthreattohumansurvival.Fromthisterriblenotioninthemindsoftoday’smostintelligentpeopleIhavetoseriouslyconsiderit.Despiteadvancesinartificialintelligenceandroboticswillbringbenefitstoourworld,butfurthercognitiveabilityistheirconcern.Thebestinterestsoftheself-awarerobotthoughtlikelytoruncountertoourhumanthinking.Thesecleverminddoes1notseemtothinkthatthe‘Terminator’isjustacrazybraingeeksholeexperiment.Yes,autonomousvehicleswillonedaybeanaccidentcausedbyhumandeath,butnotbecauseit’sself-consciousnesswouldliketodoso.Insteadsensorfailureordefectalgorithm.Ifyouputasideconcreteresults,wewillfinditandacomputerprogramstopsrespondingandnoessentialdifference.Therisksitbringsclearlydoesnotleadtohumanextinctionsuchdisastrousconsequences.Tobetterunderstandthis,wemustbetterunderstandhowmodernartificialintelligencesystemtowork.Eachapplicationcanbeusedasanoptimizationproblem,ourgoalistofindtheleastcostandmaximumreturnonthesolution.Thisartificialintelligenceisnomagic,noneedtohaveemotions.AIinsolvingtheseproblemsandourhumanwayofthinkingisverysimilar.Forexample,acardrivenbyartificialintelligencewillchoosetheshortestroutetothedestinationroute,traveling,acomputersystemwillbeprovidedfromthe2livecamerapictureofacatfoundinmiddleoftheroadanddecidedtoavoidit.Peoplealsowanttodoso.Eveninhasneverbeensopopular‘deeplearning’inthefield,thedataisplacedinneuralnetworks,argumentativeitsoperatingmodeisstilllookingfortheoptimalsolution,whichistofindandtrainwiththebestfitofthedatamodelʱ??Ajudgmentimagecontentisnotmorethanachessprogrammoreself-awareprogram.Theseapplicationsarenotreally‘think’theyarejustlookingforagivenoptimalsolution.Evensomeofthesystem’scomplexitybeyondimagination,artificialintelligencesystemsprogrammerstillonlygiventothoselinearoperation.Althoughitsoundsveryscaryrobotkillhumans,automaticvehiclecausingtheaccidentisquitescary,inacaraccidentcausedbyhumanbeingseverydaytoomany.Ofcourse,thethoughtofthesecarsarefromarealhumandrivewemighthaveahintofrelief,butstatisticallyspeaking,thesafetyofhumanbeingswillnevermatchthedriver’sdrivingautomatedroadtransportsystem.Benefitsofautonomousvehicles3beganmorethanasmallprobabilityofanaccidentevent,startingwithhighefficiencyimpactofoururbanplanning,dailytravel,transportationprice,wewillbecomemorefamiliarwithit,eventuallyautonomousvehicleslikethePCaswhenItbecameacommonpartoflife.Therealproblemwithrobotsandartificialintelligencewasbroughtinwhenitfallsintothewronghands.Likethedifferencebetweennuclearpowerplantsandnuclearweaponsthesame.Robotics,artificialintelligenceadvancesgivemaliciousbringnewtools,theycanhacktheseintelligentsystemstohurtothers.Butweshouldbeabletocontroltheserisks,andalwaysreadytofacethesechallenges.Donotforgetthathumanintelligencehasbroughtenormoussocialprogress.Gates,Musk,Hawkingremindedusofthedirectionofdevelopmentofartificialintelligenceisnotwrong.Butthegreatestharmcomesnotfromamachine,andmorelikelytocomefromtheheart.Thereisnothingmoreviciousthanthis.Wecannotimaginethetechnologywasnowbecomewithinreach,butsomeideasbutitwillalwaysremaininthecomicbook,suchasaself-conscious4humanityrobotkiller.(Source:ZDNettext/constantcompile:)5
